Website recommending technology refers to a technology of automatically recommending relevant websites upon inputting a part of a website address or description of a website by a user. This technology is applied in application scenes such as browser address bars, requiring to prompt websites the user might want to access as much as possible, or even to recommend those unknown but most appropriate to the user's demand.
At present, all the common browsers are provided with a function of recommending websites in the website bar. However, the data of the recommended websites in most browsers is from local cookies. This method has two problems in that, the websites which are not in cookies could not be recommended; and the cost for updating a website library is so high that the library could not be updated frequently. Since contents on the Internet change very frequently, it is difficult for this method to ensure the coverage and freshness of the website library.
A few browsers are able to support the function of recommending cloud websites in addition to recommending based on the local website library, and to query websites from cloud side in real time when the user is inputting and recommend a corresponding result. Compared with the recommendation based on the local website library, this method can ensure the coverage and accuracy of recommending websites. However, the existing browsers supporting this function are on the basis of a universal search engine, which uses a part of the website address or website descriptions input by the user to search in the universal search engine and recommends several top websites in the search result. Since target data of the universal search engine is web content which satisfies a requirement of the user on the query of text information, the computational algorithms of result relevance and result sorting are both designed for full-text query. But the target data of recommended websites in the browser website bar is website data, which satisfies a requirement of the user on quickly seeking websites. Thus, it is not appropriate to directly use the search results of search engines as the recommended websites, there is a problem of poor relevance and low quality, the efficiency of user in finding appropriate websites during use is low and the user's experience is not good.